USE OF PRINCIPAL-COMPONENT, CORRELATION, AND STEPWISE MULTIPLE-REGRESSION ANALYSES TO INVESTIGATE SELECTED PHYSICAL AND HYDRAULIC-PROPERTIES OF CARBONATE-ROCK AQUIFERS

Correlation analysis in conjunction with principal-component and multi ple-regression analyses were applied to laboratory chemical and petrog raphic data to assess the usefulness of these techniques in evaluating selected physical and hydraulic properties of carbonate-rock aquifers in central Pennsylvania. Correlation and principal-component analyses were used to establish relations and associations among variables, to determine dimensions of property variation of samples, and to filter the variables containing similar information. Principal-component and correlation analyses showed that porosity is related to other measured variables and that permeability is most related to porosity and grain size. Four principal components are found to be significant in explai ning the variance of data. Stepwise multiple-regression analysis was u sed to see how well the measured variables could predict porosity and (or) permeability for this suite of rocks. The variation in permeabili ty and porosity is not totally predicted by the other variables, but t he regression is significant at the 5% significance level.